http://www.translational-medicine.com/content/pdf/1479-5876-11-262.pdf

R E S E A R CH Open Access
Comprehensive analysis of the percentage of
surface receptors and cytotoxic granules positive
natural killer cells in patients with pancreatic
cancer, gastric cancer, and colorectal cancer
Yun-Peng Peng
1,2†
, Yi Zhu
1,2†
, Jing-Jing Zhang
1,2†
, Ze-Kuan Xu
1,2
, Zhu-Yin Qian
1,2
, Cun-Cai Dai
1,2
, Kui-Rong Jiang
1,2
,
Jun-Li Wu
1,2
, Wen-Tao Gao
1,2
, Qiang Li
1,2
, Qing Du
1,2
and Yi Miao
1,2*
Abstract
Background: Digestive malignancies, especially pancreatic cancer (PC), gastric cancer (GC), and colorectal cancer
(CRC), still occur at persistently high rates, and disease progression in these cancers has been associated with tumor
immunosurveillance escape. Natural killer (NK) cell dysfunction may be responsible for this phenomenon, however,
the exact relationship between tumor immunosurveillance escape in digestive malignancies and NK cell
dysfunction remains unclear.
Methods: Percentage of the surface receptors NKG2A, KIR3DL1, NKG2D, NKp30, NKp44, NKp46, and DNAM-1, as
well as the cytotoxic granules perforin and granzyme B positive NK cells were determined in patients with
pancreatic cancer (n = 31), gastric cancer (n = 31), and CRC (n = 32) prior to surgery and healthy controls (n = 31) by
multicolor flow cytometry. Independent t-tests or Mann-Whitney U-tests were used to compare the differences
between the patient and healthy control groups, as well as the differences between patients with different
pathologic features of cancer.
Results: Percentage of NKG2D, NKp30, NKp46, and perforin positive NK cells was significantly down-regulated in
patients with PC compared to healthy controls, as well as GC and CRC; reduced levels of these molecules was
associated with indicators of disease progression in each malignancy (such as histological grade, depth of invasion,
lymph node metastasis). On the contrary, percentage of KIR3DL1 positive NK cells was significantly increased in
patients with PC, as well as GC and CRC, but was not associated with any indicators of disease progression.
Conclusions: Altered percentage of surface receptors and cytotoxic granules positive NK cells may play a vital role
in tumor immunosurveillance escape by inducing NK cell dysfunction in patients with PC, GC, and CRC.
Keywords: Cytotoxic granules, Digestive malignancies, NK cells, Surface receptors
* Correspondence: [email protected]
†
Equal contributors
1
Department of General Surgery, The first Affiliated Hospital of Nanjing
Medical University, 300 Guangzhou Road, Nanjing 210029, People’s Republic
of China
2
Jiangsu Province Academy of Clinical Medicine, Institute of Tumor Biology,
300 Guangzhou Road, Nanjing 210029, People’s Republic of China
© 2013 Peng et al.; licensee BioMed Central Ltd. This is an open access article distributed under the terms of the Creative
Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and
reproduction in any medium, provided the original work is properly cited.
Peng et al. Journal of Translational Medicine 2013, 11:262
http://www.translational-medicine.com/content/11/1/262

Background
Pancreatic cancer, gastric cancer and colorectal cancer are
the most common digestive malignancies and have rela-
tively high incidences. Pancreatic cancer is characterized
by a low rate of early diagnosis and many tumors are
unresectable [1], with a 5-year survival rate of only 6% [2]
leading to a persistently high rate of mortality [3]. Colo-
rectal cancer and gastric cancer are the third and fourth
most common cancers worldwide, respectively, and are
among the leading causes of cancer-related deaths [1]. In
humans, the progression of certain malignancies is asso-
ciated with the immune function of certain lymphocytes,
such as natural killer (NK) cells. NK cells are CD16- and/
or CD56-positive, and represent the first line of immune
defense against transformed malignant cells [4].
When infection or malignancy occur, circulating NK
cells become activated by cytokines and infiltrate into the
affected tissues containing pathogen-infected or trans-
formed malignant cells [5]. The direct cytotoxic effects of
NK cells are determined by their expression of surface re-
ceptors and cytotoxic granules. NK cell dysfunction is ob-
served in patients with certain types of cancer; therefore,
surface receptors and cytotoxic granules are an important
area of cancer research.
The natural cytotoxicity receptors (NCRs) NKp30,
NKp44, and NKp46 are expressed on NK cells, as well as
T cells and NK-like cells [6-9], and mediate NK cell acti-
vation during the process of natural cytotoxicity. Killer cell
lectin-like receptor subfamily K, member 1 (NKG2D),
a C-type lectin-like protein, is an activating receptor
expressed on NK cells and also gamma-delta T cells,
natural killer T (NKT) cells and other types of immune
cells [10]. NKG2D is required for the ability of NK cells to
lyse harmful cells [11,12]. NK cells also express other acti-
vating receptors including DNAX accessory molecule-1
(DNAM-1) which binds to two well-characterized ligands
(nectin-2 and the poliovirus receptor) and exerts similar
effects to NKG2D [13]. Killer cell lectin-like receptor sub-
family C, member 1 (CD94/NKG2A-B), killer cell lectin-
like receptor subfamily C, member 2 (CD94/NKG2C-E)
and the killer immunoglobulin-like receptors (KIRs) are
described as inhibitory receptors, which are important for
the education of NK cells and NK-induced cytotoxicity
through interacting with the major histocompatibility
complex (MHC) class I allotype [14]. The cytotoxic gran-
ules perforin and granzyme B are intracellular molecules
present in a number of lymphocytes, including NK cells.
Perforin is required for the ability of granzyme B to pro-
mote apoptosis in target cells [15,16]. NK cells express
high levels of perforin and granzyme B, and the expression
levels of these molecules are closely associated with the
cytotoxicity of NK cells [17].
Methods
Patients and healthy controls
Patients diagnosed with PC (n=31), GC (n=31), or CRC
(n= 32) who were treated at Jiangsu Province Hospital
were enrolled in this study. All patients had only received
positive preoperative preparation and had not undergone
radiotherapy, chemotherapy or any other therapeutic
strategies prior to surgery. The main clinicopathological
features of the patient cohorts are shown in Table 1. All
peripheral blood samples were collected from the patients
before surgery, and peripheral blood samples from 31
healthy control individuals were provided by Jiangsu
Province Blood Center. This study was approved by
the Ethics Committee of the First Affiliated Hospital of
Nanjing Medical University. Each of the patients and
healthy control individuals gave informed consent.
Reagents
The anti-human CD3-FITC/CD16 + 56-PE mixed antibody
wasobtainedfromBeckmanCoulter(Brea,CA,USA).
The anti-human CD3-FITC, CD16-PE/Cy7, CD56-PE/Cy7,
Table 1 Clinicopathological features of the patients and healthy controls included in this study
Clinicopathological characteristics
Groups Healthy controls Pancreatic cancer Gastric cancer Colorectal cancer
n=31 n=31 n=31 n=32
Gender Male 21(67.7%) 23(74.2%) 20(64.5%) 17(53.1%)
Female 10(32.3%) 8(26.8%) 11(35.5%) 15(46.9%)
Age Median age 53 64 61 60
Range 35-57 34-76 35-82 40-82
AJCC Stage
*
0 0 1(3.2%) 0
I 0 7(22.6%) 1(3.1%)
II 20(64.5%) 4(12.9%) 16(50.0%)
III 0 17(54.8%) 14(43.8%)
IV 11(35.5%) 2(6.5%) 1(3.1%)
*
2010 American Joint Committee on Cancer (AJCC).
Peng et al. Journal of Translational Medicine 2013, 11:262 Page 2 of 10
http://www.translational-medicine.com/content/11/1/262

NKG2D-PE/Cy7, NKp44-APC, NKp46-PE/Cy7, NKp30-
APC, KIR3DL1-PE, DNAM-1-Alexa Fluor 647, and
perforin-PerCP/Cy5.5 antibodies, and the RBC Lysis Buf-
fer, Fixation Buffer and Wash Buffer were purchased from
Biolegend (San Diego, CA, USA), as well as FITC, PE, PE/
Cy7, APC, PerCP, Alexa Fluor-647, and PerCP/Cy 5.5
mouse IgG1 antibodies. The anti-human NKG2A-PerCP
and granzyme B-APC antibodies were obtained from
R&D Systems (Minneapolis, MI, UAS). All antibodies
were mouse monoclonal antibodies.
Preparation of peripheral blood samples and flow
cytometric analysis
Each peripheral blood sample (2 ml) was aliquoted into
four tubes (100 μl per tube), which were labeled tube-1,
tube-2, tube-3 and tube-4, respectively.
Peripheral blood samples of tube-1, tube-2 and tube-3
were stained to detect surface receptors as follows. Firstly,
to identify NK cells, anti-human CD3-FITC/CD16 + 56-
PE mixed antibodies were added to tube-1 and tube-2.
Anti-human CD3-FITC, CD16-PE/Cy7 and CD56-PE/Cy7
antibodies were added to tube-3. Secondly, anti-human
NKG2D-PE/Cy7 and NKp44-APC antibodies were added
to tube-1. Anti-human NKG2A-PerCP, NKp46-PE/Cy7
and NKp30-APC antibodies were added to tube-2. Anti-
human KIR3DL1-PE and DNAM-1-Alexa Fluor-647 anti-
bodies were added to tube-3. The three tubes were
incubated in the dark at room temperature for 15-20 min.
Then 2 ml RBC Lysis Buffer was added per tube. After
incubating in the dark at room temperature for 15 min,
the cells were washed twice with PBS.
Peripheral blood sample of Tube-4 was stained to de-
tect cytotoxic granules as follows. Firstly, anti-human
CD3-FITC/CD16 + 56-PE mixed antibodies were added
to tube-4 to identify NK cells. After incubating in the
dark at room temperature for 15-20 min, 2 ml RBC
Lysis Buffer was added per tube, and the mixtures were
incubated in the dark at room temperature for 15 min.
Then the cells were washed twice with PBS and fixation
Buffer (500 μl per tube) was added. The mixtures were
incubated in the dark at room temperature for 20 min,
and then the cells were washed twice with Wash Buffer.
Lastly, anti-human perforin-PerCP/Cy5.5 and granzyme
B-APC antibodies were added to tube-3. After incuba-
ting in the dark at room temperature for 15 min, the
cells were washed twice with PBS.
Flow cytometric analysis
According to cell physical characteristics, forward scatter
(FSC) and side scatter (SSC), a cell subset located in left
lower quadrant (PBMCs) was selected from total cell
subset and defined as gating “A”. And then, according to
cells staining, another cell subset which detected as
CD3-/CD(16 + 56) + (NK cells) was selected from gating
“A”and defined as gating “Q”. Further detections for
surface receptors and cytotoxic granules were based on
cells from gating “Q”. The whole detection for per tube
would stop until getting 10000 cells from gating “Q”.
Isotype control was applied in our study to exclude non-
specific fluorescence using matched isotype monoclonal
antibodies (FITC, PE, PE/Cy7, and APC mouse IgG1
antibodies for tube 1; FITC, PE, PE/Cy7, PerCP, and
APC mouse IgG1 antibodies for tube 2; FITC, PE, PE/
Cy7, and Alexa Fluor-647 antibodies for tube 3; FITC,
PE, PerCP/Cy 5.5, and APC mouse IgG1 antibodies for
tube 4). Data were detected by multicolor flow cytome-
try (Gallios, Beckman Coulter, Brea, CA, USA) and
gallios software (Beckman Coulter, Brea, CA, USA), and
analyzed by Kaluza software (Beckman Coulter, Brea,
CA, USA).
Statistical analysis
Independent t-tests were used to compare the differences
between two groups when the two groups both accorded
with normal distribution, otherwise Mann–Whitney U-
tests were used. Independent t-tests and Mann–Whitney
U-tests were performed using Statistical Product and
Service Solutions 19.0 (SPSS 19.0) (SPSS Inc., Chicago, IL,
USA). Data were expressed as means ± standard devia-
tions (Mean ± SD). The level of statistical significance ac-
cepted was P < 0.05.
Results
Percentage of surface receptor and cytotoxic granule
positive circulating NK cells
We determined the percentage of seven surface receptors
positive circulating NK cells in both healthy controls and
patients with PC, GC, and CRC by multicolor flow cyto-
metry. The percentage of tested molecules positive circu-
lating NK cells of the cancer patients and healthy controls
are presented in Figure 1 and Table 2.
Compared to the healthy controls, significantly decreased
levels of activating receptors NKG2D, NKp30, NKp46,
and DNAM-1 positive NK cells were observed in PC
patients (P<0.001, P<0.001, P<0.001, and P< 0.01, re-
spectively); however, an significantly increased level of
inhibitory receptor KIR3DL1 positive NK cells was ob-
served in patients with PC (P< 0.001). In GC patients, the
activating receptors NKG2D, NKp30, and NKp46 positive
NK cells were also significantly down-regulated compared
to the healthy controls (P<0.001,P<0.001,andP<0.001,
respectively); however, the inhibitory receptor KIR3DL1
positive NK cells was also significantly up-regulated com-
pared to the healthy controls (P< 0.001). Furthermore, the
levels of activating receptors NKG2D, NKp30, and NKp46
positive NK cells in CRC patients was significantly lower
compared to healthy controls (P<0.01, P<0.001, and
P< 0.001, respectively); however, the level of inhibitory
Peng et al. Journal of Translational Medicine 2013, 11:262 Page 3 of 10
http://www.translational-medicine.com/content/11/1/262

receptor KIR3DL1 positive NK cells was also significantly
higher compared to the healthy controls (P<0.001).
We also determined the percentage of cytotoxic per-
forin and granzyme B positive circulating NK cells in
both healthy controls and patients with PC, GC, and
CRC (Figure 1 and Table 2). Respectively compared to
the healthy controls, percentage of perforin positive NK
cells was significantly lower in patients with PC, GC,
and CRC (P< 0.01, P< 0.001, and P< 0.001, respectively).
Percentage of granzyme B positive NK cells was at high
levels in both the NK cells of the patients with cancer
and the healthy controls.
Altered percentage of NKG2D, NKp30, NKp46, KIR3DL1,
and perforin positive NK cells correlate with disease
progression
The correlations between the percentage of NKG2D,
NKp30, NKp46, KIR3DL1, and perforin positive NK cells
and the pathologic features of PC, GC, and CRC are res-
pectively shown in Tables 3, 4 and 5.
In pancreatic cancer, NKG2D, NKp30, NKp46, KIR3DL1,
and perforin had no association with the presence of
distant metastasis. In non-metastatic pancreatic cancer,
the percentage of NKG2D and NKp30 positive NK cells
were significantly decreased in patients with lymph
Figure 1 Distribution of the percentage of surface receptor and cytotoxic granule positive circulating NK cells in healthy controls
(Controls) and patients with pancreatic cancer (PC), gastric cancer (GC), and colorectal cancer (CRC). The bar for each group was means
and standard deviations (Mean and SD).
Peng et al. Journal of Translational Medicine 2013, 11:262 Page 4 of 10
http://www.translational-medicine.com/content/11/1/262

node metastasis than patients without lymph node
metastasis (both P< 0.05). The levels of NKG2D and
perforin positive NK cells were significantly lower in
patients with blood vessel invasion, compared to pa-
tients with non-metastatic pancreatic cancer who did
not have blood vessel invasion (P<0.05 and P<0.01).
NKp46 positive NK cells percentage also correlated
closely with the histological grade in non-metastatic
pancreatic cancer (P<0.01).
In gastric cancer, the percentage of NKG2D, NKp30,
and perforin positive NK cells were significantly lower in
patients with lymph node metastasis than patients without
lymph node metastasis (P<0.01, P<0.05 and P<0.05,
respectively). NKG2D positive NK cells were signi-
ficantly down-regulated in patients with blood vessel
invasion compared to patients without blood vessel in-
vasion (P< 0.05). NKG2D, NKp30, and perforin positive
NK cells were significantly higher levels in patients with
gastric cancer who had well or moderately differentiated
tumors, compared to those with poorly differentiated
tumors (P<0.01, P< 0.05, and P< 0.05, respectively).
Moreover, the percentage of NKp30 positive NK cells
correlated significantly with the depth of invasion in
gastric cancer (P<0.05).
In colorectal cancer, NKG2D, NKp46, and perforin
positive NK cells were significantly lower levels in
patients with lymph node metastasis compared to
patients without lymph node metastasis (P<0.01,P<0.05,
and P< 0.01). The percentage of NKp30, NKp46, and
perforin positive NK cells correlated markedly with
depth of invasion in CRC (all P<0.05). The percent-
age of NKG2D and perforin positive NK cells corre-
lated closely with histological grade in CRC (P<0.01
and P< 0.05). None of the molecules tested were
associated with blood vessel invasion or nerve invasion
in CRC.
Discussion
In this study, we quantified the percentage of several
activating and inhibitory surface receptors positive
circulating NK cells, as well as the cytotoxic granules
perforin and granzyme B, in patients with PC, GC,
and CRC. The balance between activating and inhibi-
tory receptors has been shown to be a key factor
which determines NK cell activity [18]. It has been
demonstrated that NK-mediated anti-tumor immunity
is frequently defective in patients with certain malig-
nancies [19,20]. This study indicates that patients
with PC, GC, and CRC have dysfunctional NK cells;
therefore, NK cell dysfunction may be an important
component of tumor escape from immunosurveillance in
these cancers.
NKp30, NKp44, and NKp46 are the most well charac-
terized NCRs. Our results show for the first time that
the numbers of NKp30 and NKp46-positive NK cells
were significantly reduced in almost all patients with PC,
GC, and CRC, consistent with studies in other malig-
nancies such as cervical cancer, breast cancer, and mela-
noma [21-23]. It has been reported that NCR-positive
NK cells have the ability to kill harmful cells, such as
transformed malignant cells and infected cells, and can
also secret inflammatory cytokines such as interferon-γ
(IFN-γ) and tumor necrosis factor-α(TNF-α) [24]. Ac-
cordingly, the lower number of cells expressing NKp30
and NKp46 may be partly responsible for the poor func-
tion of NK cells in patients with PC, GC, and CRC. The
NCR-mediated interaction between NK cells and their
target cells is ligand-dependent. Cellular heparin or
heparin sulfate proteoglycans, which are expressed at
high levels on cancer cells, are ligands for all NCRs [25],
while natural killer cell cytotoxicity receptor 3 ligand 1
(B7-H6) and BCL2-associated athanogene 6 (BAT3) are
specific ligands for NKp30 [26,27]. The specific ligands
Table 2 Respective comparison of the percentage of surface receptors and cytotoxic granules positive NK cells in
healthy controls and three kinds of cancer patients
Healthy controls Pancreatic cancer Gastric cancer Colorectal cancer
%%P%P%P
NKG2A 1.5 ± 1.0 1.9 ± 1.0 ns
U
1.3 ± 1.2 ns
U
1.4 ± 1.0 ns
U
NKG2D 92.1 ± 2.7 83.4 ± 8.4 < 0.001
T
86.9 ± 2.9 < 0.001
T
84.9 ± 10.0 < 0.01
U
NKp30 71.8 ± 15.3 46.5 ± 20.2 < 0.001
U
36.6 ± 21.3 < 0.001
U
42.6 ± 19.4 < 0.001
U
NKp44 1.8 ± 0.9 2.4 ± 1.7 ns
U
1.7 ± 1.4 ns
U
0.9 ± 1.3 ns
U
NKp46 84.8 ± 5.1 56.4 ± 22.2 < 0.001
U
50.1 ± 29.5 < 0.001
U
55.1 ± 24.0 < 0.001
U
DNAM-1 91.4 ± 4.6 87.2 ± 5.5 < 0.01
U
87.7 ± 7.8 ns
U
89.5 ± 5.1 ns
U
KIR3DL1 9.6 ± 2.7 19.8 ± 11.4 < 0.001
U
17.9 ± 12.3 < 0.001
U
19.5 ± 13.1 < 0.001
U
Granzyme B 97.9 ± 1.6 96.9 ± 2.4 ns
U
97.8 ± 2.0 ns
U
97.4 ± 2.5 ns
U
Perforin 95.2 ± 3.0 79.9 ± 16.0 < 0.01
U
81.2 ± 15.5 < 0.001
U
76.3 ± 19.1 < 0.001
U
U
represented Mann–Whitney U-tests and
T
represented independent t-tests. Data were expressed as means ± standard deviations (Mean ± SD).
Peng et al. Journal of Translational Medicine 2013, 11:262 Page 5 of 10
http://www.translational-medicine.com/content/11/1/262